The idea that some day computing functionality will become a utility like electricity, delivered through a plug in the wall from central plants run by computing companies and paid for according to how much you use them, is not new. Actually it appears in science fiction from the 1950s. The difference today is that anyone who looks at current trends can see when that will happen – and the answer is now.
The utility model has a lot of benefits for users. But for those who work in the industry it means wrenching change, equivalent to what happened to the power industry in the 1920s. If the organizations many of us work for are going to buy increasing amounts of their IT from utility providers – traditional outsourcers, Software-as-a-Service companies, “The Cloud” -- then the IT organizations in those companies are going to shrink and eventually disappear. In fact, they have been doing just that gradually since about 1990, and the idea that IT today offers huge career opportunities is a myth based on a past that has already ended. And for many of us, both in IT organizations and those that serve them, such as consultancies and the large analysis organizations (where I have worked for years), this means our careers are on life support.
So why do I say this? Let's look at the trends.
At about 1990 packaged software first started appearing for large computer systems (above the desktop). At that time the largest part of most IT organizations was the programming group. It wrote, updated, and maintained the big mainframe applications on which businesses ran. At the time I was working for Gartner Group, and our clients took the view that packaged software would only work for “well defined applications.” So today where are all those legions of programmers? Pretty much gone. I have a lifelong personal friend, for instance, who was a top Fortran programmer on Wall Street. One morning he walked into work to be told that he, and most of the people he worked with, were no longer needed. Today he lives in Florida and works as a real estate appraiser and, during tax season, a supervisor for H. & R. Block. Most of what custom programming is being done has moved to India and is delivered via the Internet, but most of the software running in corporate IT data centers are packaged apps. Oracle and SAP have built empires by supplying highly customizable packaged software for applications that definitely are anything but “well defined” and that evolve rapidly over time.
Another example is computer operations. Of course data centers and desktops will always require some people on site to do physical tasks (at least until robots replace them). But in fact today many data centers run “lights out”, with no permanent local staff, and with all activities from tape mounting through software updating to complete restarts either automated or controlled remotely, through the Internet, often from overseas. Delivered out of a plug in the wall and paid for as they are used.
In the last few years we have witnessed two further steps toward the end of IT as we know it – Software-as-a-Service (SaaS) and cloud computing. This story starts early in the new century. At that time the “next big thing” was customer relationship management (CRM) and the top vendor was Sieble Systems. Then, at about 2005, Salesforce.com appeared. Suddenly instead of buying Sieble software, and hardware to run it on, and hiring a crew of expensive IT experts to tend it, companies were quite literally getting their CRM out of a plug in the wall and paying for what they used. Today we hear IT experts say that SaaS will only work for “well-defined applications”. But while Salesforce.com remains the big SaaS name, in fact dozens, maybe hundreds of SaaS companies have appeared in the last five years, among them covering a tremendous amount of business IT, and almost all of them are growing daily. This industry is growing so fast and is so large already that a boutique analyst company specializes in following them. Those who want to know more about SaaS should contact Saugatuck Technology (www.saugatech.com).
Cloud computing is an invention of Google and other Internet-based companies. Today it is focused more on providing storage and other generic services to individuals and very small organizations, but anyone who thinks they aren't targeting large enterprises is living in a fantasy world. The day will come, and probably soon, when most data archiving and an increasing amount of active data handling will move out of the data center and into the cloud.
So why is this inevitable? Because it makes excellent business sense. Unless you work for a vendor, the focus of your enterprise is not computing. That, like electricity, is simply something required to do business as far as your senior management is concerned. If the enterprise can get a business advantage by leveraging advanced computing, as Wal*Mart famously did in the 1980s and '90s, then it becomes strategic. But inevitably last year's source of competitive advantage becomes next year's cost of doing business as all your competitors either catch up or disappear. Once that happens it becomes much more efficient to buy that computing functionality from a SaaS vendor than to keep it in house, and if your organization is one of the followers it never makes sense to do it in house. Inevitably, leveraging SaaS and other types of computing service providers will itself become a source of business advantage for early adapters simply because it is cheaper. The supplier can afford the best and most complete staff, the most efficient equipment, etc.
Back in the 19th Century the sign of industry and progress was the belching smokestack. Every company, every factory, regardless of whether it made cloth or automobiles, had a power department with a mechanical engineer as a boss, mechanics who kept the machines running and installed new ones when needed, and a lot of people shoveling coal into furnaces. Late in the century electricity replaced mechanical power, which meant that electrical engineers and electricians took over the team, but those smoke stacks were still there. Then roughly around 1900 a couple of very smart people named Edison and Tesla figured out how to generate electricity centrally and send it long distance through wires. By the 1920s increasing numbers of companies were buying power from utilities rather than generating it themselves because it was cheaper. Then 1929 hit. Companies either had to get lean and efficient or go under, and those power departments were obvious targets. When was the last time you saw belching smokestacks at a company headquarters?
Today we are in an economy that at least in the United States is being compared to 1929. Companies need to get much more efficient or face bankruptcy.
Action Item: Take a hard look at your career, your skills, and the trends in IT employment at your company, and ask yourself, “What will I be doing in five or ten years?” Unless the answer is “working for a vendor” or “retired” you might want to start planning for life beyond IT, like my friend in Florida.