This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.
Justifying the Cloud
What separates a critical technology trend from extravagant media hype? The answer is "implementation," and the first critical step in that direction is justification. For every enterprise expense there has to be a value proposition, a hope of ROI. In troubled economic times, enterprises raise their ROI targets to factor in their additional economic risk, which makes justification all the more important.So when we look at cloud computing, the big question for 2010 won't be the technology or vendor choices, it will be the justification. Can cloud computing meet the higher expectations of a still-tentative recovery? Not in the usual way, it turns out, but perhaps by thinking outside the box.
The classical justification for cloud computing is reduction in costs. Whether you're talking about private or public clouds, the core value proposition is resource economy of scale. Virtualization saves server dollars by letting enterprises share servers among applications. Cloud computing expands the base of resources you can share from a single server or local data center to potentially a whole world of servers. That can reduce server cost for private clouds or create economies in public clouds that let services from the cloud beat in-house IT costs.
The problem with this approach is the old CFO bugaboo of risk. To start with, any economy-of-scale concept needs to create scale to generate economy. But scale means jumping convincingly into cloud technology, which means bearing an enormous market risk, whether you're an enterprise or a provider. Thus, your first tentative steps into cloud computing are hardly likely to be world-wide commitments, and yet to generate any incremental cloud resource economy you need to expand the cloud far beyond the current data-center limits of virtualization.
Look at a hypothetical private cloud justification dialogue as an example. The CIO says "I'd like to change the structure of all our data centers, put all our critical applications into a new execution environment, and increase our dependence on our networks drastically." That's when the CFO's calculator comes out and, factoring in the risk premium, you get a triple-digit ROI target that no project on earth could possibly meet.
You could also theoretically justify substituting public cloud services for private IT investment based on cash flow management. When a business buys equipment, the dollars flow out immediately but the expenses can't be taken all at once. Instead the equipment has to be written down over three to five years. For companies with cash flow issues, that imbalance between when the money goes and when the expense can be taken can be critical. In effect, cloud computing could convert a capital cost that has to be written down over multiple years into an operating expense that can be taken immediately.
But even this isn't a compelling value proposition. Larger firms probably have millions in capital assets already, and only the new IT and network equipment that might otherwise have been needed can contribute to a cash-flow justification for cloud computing. For small and mid-sized businesses, cash flow management may be a great reason to jump into cloud computing, but it probably won't work for larger enterprises.
Cloud computing could move forward with these two traditional justifications, but not very rapidly. What's interesting, according to surveys I've been conducting among enterprise IT buyers, is that there is increased interest in a third cloud value proposition. This one makes a benefit out of one of the biggest barriers to cloud adoption-risk.
Suppose we go back to our CFO with a project request, this one to create a new collaboration system, or a sales management system, or even a new call center. There's all kinds of hypothetical benefits to be had, of course, but the gear that will be needed is an investment the company will have to pay back over years of successful use. Will all the benefits pan out? Will technology changes obsolete our whole idea before we really even get it going? The CFO calculator starts humming, and there goes the project down the ROI-target tube.
But suppose instead we say that we're going to take a one-year test drive with the new system? There's no up-front capital cost, no expensive gear to write down over years of time. Technology risk is carried by our provider and not by us. If we don't like what we find after a year, we walk away and either fall back to an older approach or try another new one. Sure, there are costs associated with transitioning to (and, at worst, from) the new system, but they're a lot lower than the costs of tossing out everything we buy and starting over. Could this bring a smile to the face of the CFO?
In January, the number of enterprises who'd even thought of this justification was in the statistical noise level. In October, 38 of 78 enterprises who were looking at a significant cloud commitment in 2010 thought this risk-managing approach was the best hope of justification.
Project failures, defined as situations where the project fails to meet its ROI targets, have climbed in the last two years to nearly 40% of all projects, according to my 277-enterprise survey base. Companies like HP and IBM are seeing professional services as a larger piece of their total opportunity, because the complexity of technology is getting beyond what enterprises are able to comfortably apply to their business needs. The best thing that the cloud might do for us is manage that risk.
Another interesting truth here is that the enterprises that saw risk-based justification of cloud projects as critical were three times as likely to want software- or platform-as-a-service versus infrastructure-as-a-service. They want the cloud to do more, perhaps even do it all. Risk-based cloud justification is the only cloud computing value proposition that drives the enterprise into large-scale IT outsourcing eventually. That could mean that 2010 could be the start of a profound change in enterprise IT.