Admittedly, this hasn’t been the best week for the reputation of cloud computing. Recent outages have resurrected arguments about the wisdom of moving critical functions to the cloud. But those who would use this as a reason to dismiss the cloud altogether need to put this situation into perspective. 

First, for salesforce.com and many other services, the cloud is not merely a method of delivery. Salesforce.com performs a number of functions that can only take place in the cloud. In particular, for a globally dispersed organization, the ability to collaborate in real time is something that requires web deployment, and can’t really be duplicated with on-premise software. As business becomes more globally-focused, such capabilities will become increasingly more important. 

Secondly, the cloud makes it possible for some companies to develop a robust IT infrastructure that they would otherwise not be able to afford. This is particularly true for small and medium-sized businesses, for whom in many cases it’s not a question of Cloud vs. on-premise, but rather one of cloud vs. no IT infrastructure (or at best a vastly inferior infrastructure).

The advantages of cloud-based applications go way beyond cost savings. Companies can adopt, iterate, and drop technologies very quickly in response to rapidly changing business needs. This is not just a convenient side benefit -- as the business cycle continues to grow more rapid and erratic, the ability to adapt very quickly is becoming a quality that companies need to survive. On-premise applications sometimes require years to implement; depending on function, many who insist on going on-premise all the way will find applications outdated before they are rolled out. 

Additionally, keep in mind that application disruptions occur with on-premise software too. It’s not like this is a cloud-only phenomenon. Sure, when it happens in-house it’s easier to pin the blame on someone and punish them. But that doesn’t get your application running any faster. Being “at the mercy” of vendors is a matter of perspective. Vendors like Salesforce, Google and AWS have tremendous resources and every incentive in the world to make such outages as quick and painless as possible. 

When you’re relying entirely on internal capabilities, you’re at the mercy of your own budget constraints and the decisions you have made earlier on in the year. If you’ve skimped on your IT budget, application disruptions may be long and painful. For that matter, they may be painful even if you’ve invested heavily in IT resources, because you can’t scale with on-premise like you can in the cloud. 

In the end, what does this mean for the cloud? Probably not much more than the first round of power outages meant for the electrical grid. WSJ’ Michael Hickins provided a sane perspective, stating that the recent criticism “won’t change the basic economics of cloud computing, which allow businesses to shift capital expenditures and labor costs to third-party vendors, and to reallocate their IT resources to activities that create more value. But the outages will put pressure on vendors to provide customers with more up-to-the-minute updates about outages, their causes and remediation, and most importantly, estimates of when issues will be resolved.”