The discussion of if the "Cloud" is a good idea, or if we are "going to the cloud" has been replaced with how we are going to the cloud and what if anything should stay in house. The fact is that doing otherwise is rapidly becoming unsustainable.
In this second decade of the 21st century, Enterprise software has a challenging agenda:
It must be robust, scalable, and secure. It must be consumer usable, while being customizable. It must be easy to evolve, change and upgrade. It must be easy to integrate its data, services, UI and identity management with legacy, current and future systems. It must, of course, enable the specific processes, transactions, communications and other manipulations that are unique to your organization. It must be cost effective.
None of these requirements is negotiable — but some companies are making surprising tradeoffs, because the payoff is vast. The good news is that the issues that concern most organizations most deeply — security, robustness and records management are rapidly resolving. Of course, consumer cloud applications are not the same thing as enterprise cloud applications — there are different standards for reliability, recoverability and security. One should not be confused for the other. But the race is on. Enterprise companies are racing to deliver applications that are both usable and cloud-deployed on top their robust backends, and consumer cloud companies are picking at the edge of enterprise.
Why are we making tradeoffs? Many organizations, including the US Government and giants like GM, who are rumored to be going over to Google from IBM, are beginning a move to the cloud. While the cloud has some intrinsic benefits — like universal access, integration, standards, social and the like, the real reason we value the cloud isn't so much that the cloud is inherently wonderful, but that the alternative is strangling enterprise. The staggering complexity, expense and time for an IT organization to meet the needs of even a moderately sized organization is becoming unsupportable, and beginning to hold good business back.
The basic value and issues of cloud computing were recently, and perhaps canonically described by Andrew McAfee. But at its core, the reason the cloud debate is extremely simple: SaaS lets you get the best available tools faster, cheaper and with a lot less planning and decision making. Your users will be happier, and your IT focus can focus on 1. that which is unique to your company; 2. monitoring and curating technology usage and adoption in the enterprise; and 3. Thinking strategically about where to go next.
1. Find Focus:
You might have the best IT staff in the known universe. And then again you may not. It is immaterial. No matter how hard you work, how rich or how smart you are, Joy's Law prevails: "No matter who you are, most of the smartest people work for someone else”. [or for the more mathematically geeky, I like this formulation: "Most smart people don't work here… for any definition of here"]. This is to say that the world is filled with smart people, but you can only have a relative few of them on your team. So it is irrational to try to replace aggregate work of these hordes of competents with the few you have on your staff.
The role of IT is to help employees optimize their use of technology to enable them to best meet the needs of their market and customers at the lowest cost. The cloud gives IT the luxury of being able to meet basic needs at the same time as envisioning and shepherding the larger IT picture for the organization, without grinding itself into dust. The IT department can now curate the adoption and use of technology, rather than stand like atlas trying to just keep it aloft.
2. Make What You Can't Buy:
Maximize your Person-Bytes, or why it took Thomas Thwaite a year to make a toaster (that didn't work).
This concept overlaps with Joy's law, but it is subtly different. Not only can you not corner the market on a particular expertise, you need an ever broader swath of expertise to make progress these days. This is precisely articulated by Ricardo Hausmann, an economist who gave one of the most interesting lectures I've seen this year. He begins by comparing the relative genius of the Inuit man who can build his own house, acquire his own food and fashion his own snowshoes, with the relative helplessness of "Modern White Guy" who is, by these measures, completely useless. He goes on to describe knowledge in terms of "Person Bytes" which is a rough measure of how much knowledge an individual can master. He uses fascinating measures and statistics to show that those societies which can produce goods that require the most person-bytes, are the most differentiated, and enjoy unimpeded economic growth.