Every Project Will Fail

We all know the industry reports (particularly the CHAOS report of the Standish Group) saying that only a small number of software projects are successful. But when is a software project successful? People have been struggling to find a proper definition for project success for years, and nobody seems to have succeeded. One traditional view was that a project is successful when it is delivered on time, within budget, and according to specifications. A newer view is that a project is successful whenever it meets a customer’s expectations, whatever they happen to be at the time of delivery. Others think a project can only be successful when it has really contributed to the success of a customer’s business in a measurable way. But I think they are all wrong.

Do you think dinosaurs were successful? Do you think humans are successful? I suspect that many people answer ‘no’ to the first question and ‘yes’ to the second. However, dinosaurs have ruled the earth for about 150 millions years, while humans have only been around the block for less than 200,000 years. It appears that we need an additional 149.8 million years to prove that we are more successful than the dinosaurs. And do you think horses are successful? I think my daughter does. But the great and late paleontologist Stephen Jay Gould didn’t. He pointed out that all variations of horses have almost been eradicated from the Big Biological Tree of Life, with the exception of the domesticated breeds. The few remaining horses can only be considered successful in the sense that they have convinced humans to sit on them and to hit them with a horse whip; otherwise they would have gone extinct. I think it is apt to say that every new species is a success until it fails and goes extinct. Given the fact that 99.9% of all species are now extinct, it is evident that failure is in abundance.

It is the same with software projects. Every software project is a success, until it fails. I have known projects that were a success for only a very short time, until the customer finally figured out what he really wanted, which appeared to be something else. I have known projects that were a success (on time, within budget and according to specifications) until it appeared that they could not live up to our customers’ expectations. And I have also know projects that are still successful, many years after their very first release date. They have been able to postpone their failure by continuous adaptation and evolution. But they will fail, someday. I’m sure of that. I'll simply have to wait a while.

  • Arthur C. Clarke (or HAL vs. IBM)
  • New column in The Software Practitioner
Related Posts
free book
GET MY FREE BOOK!
“How to Change the World”