|
One of the standard responses to the Year 2000 Problem is this one: "They'll fix it." Who they are is never said. We can now safely respond, "No, they won't." (By safely, I mean safe from refutation by future developments. Our conclusion, therefore, is that we are all terribly unsafe.) The document attached to this link shows why. It was posted by the Standish Group. This report is based on extensive surveys and detailed follow-up. It is considered authoritative in the software industry.
Sofware projects come in late. The larger the project, the truer this assessment is.
COMPLETION: Over 30% of all software projects are cancelled before their completion.
TIME: Projects come in on time in about 16% of all cases. For large firms ($500m/year revenue), it is 9%.
FEATURES: Of this 9%, only 42% of the original features are delivered. Small companies ($100m to $200m) do much better: 78% of the projects get deployed with 74% of the original features.
COSTS: Completed projects will cost on average 189% of the original estimates. This does not factor in opportunity costs of any delays, which there will almost always be.The larger firms spend 178% more. Medium firms ($200m to $500m) spend 182% more. Small firms spend 214% more.
The Party Line of the American media is this: "The biggest firms have the revenues required to finish their y2k projects. Small and medium firms do not. There is greater risk that small and medium-sized firms will not make it." This Party Line is invariable when discussing banking. This line of reasoning is ridiculous. Because banking is an international system, the failure of many foreign banks -- which are way behind U.S. and Canadian banks -- will take down ours. This is never, ever discussed in the major media. But even the original thesis is faulty. Smaller is safer when it comes to software projects.
Y2K is the largest software project in history. Because of the interconnection of computers and systems, this is one gigantic repair project. The Millennium Bug is systemic. But it is not being dealt with as a system. There is no international authority capable of dealing with it as a system. There is no agreed-upon standard for what constitutes date compliance (i.e., where to place the century's numbers, and how many -- two or four). This, the media refuse to discuss. Reporters do not understand the systemic nature of the y2k problem. If they did, they would quit and move to a small town. Only those who do not understand the systemic risk remain on major media staffs. Their reporting reflects this fact.
In any case, the original premise -- "larger is safer" -- is wrong. The truth is the opposite, according to this report. Smaller firms are far more likely to complete their repairs. Of course, this is only true if they actually begin their repairs. Most organizations -- large and small -- began late or have not yet begun.
This report has been widely quoted, but it is rarely read in its entirety. In fact, the organization goes out of its way to prohibit anyone from reproducing it. But since they have posted their report, I can lawfully summarize it as a public document. I encourage you -- BEG you -- to click through, print it out, read it, highlight it, and see if I'm summarizing it accurately.
Then you had better make plans.
|