A little over three years ago a reasonably successful political campaign in the US ran on the premise that change is good. Since we try to stay particularly neutral when it comes to political commentary, we can't go as far as to agree, or disagree with that statement. Moreover, we'd much rather stick with the understanding that change is inevitable. In fact, we'd go as far as to openly suggest that change is much more an agent of cause with both positive and negative effects. Taking that a bit further, change can create, resolve, or even prevent issues, and is thus correlative to cost.
Of course, determining the true cost of change is hard, arguably almost impossible. Even if you recover remarkably, experience little down time, or attribute no dollar value with a change, there may still be inherent cost to your organization. In many cases the changes made, even those that reach resolution of an issue, may actually impair your ability to find the root cause.
To make matters worse, this equation becomes compounded when response to the effect or result of a change is handled poorly, especially with regard to timing. In a sense, we often see a directly proportional relationship of cost to timing (read: how long it takes to resolve an issue caused by change). To demonstrate this, let's take two recent examples.
First, look at RIM's outage about two months ago that spread from Europe to North America, and then around the rest of the world. Months after the issue has been "resolved," RIM can't determine the absolute cause of the issue. This means the company more than likely couldn't prevent the same issue from happening again, and at the very least doesn't have the proper mechanisms in place to make the essential changes necessary to resolve any similar issue in the future. Now, take a look at almost the same issue experienced by Amazon, and loss of service to their EC2 services. Ultimately traced back to network configuration change, it took Amazon in what now seems like lightening speed, a few days to determine the actual cause.
While all of RIM's woes can't be taken back to this single incident, there is associated cost, especially with regard their customer's view on how they handle and resolve issues. In contrast, comparing to Amazon's handling of the issue, we can extrapolate RIM's mechanisms for understanding and responding to change are lacking. However, given the availability of modern, intuitive, and cost effective tools like ChangeGear, customers will likely deem that unacceptable in the long term.