11/02/2009 02:48:26 PM EST

Interesting article. There is, additionally, another key reason for project failures: lack of viable cost & schedule estimates http://www.galorath.com/index.php/library/abstract/a-case-for-software-estimation/ in the first place along with insufficient tracking and control during development. Many say more projects fail due to lack of planning than any other reason. Hence programs end up on “death marches” http://www.yourdonreport.com/index.php/2007/11/27/death-march-presentation-download/ get canceled after huge wastes of resources, or ever provide business value sufficient to justify their cost. According to Donsani 80% of projects cost more than they return.

United StatesDan Galorath

11/03/2009 04:55:27 AM EST

It seems that every time I read an article on software project failure the writers attribute such failures to just about the same issues they have been arguing over for the past 30+ years. As a senior software engineer for 35 years I have seen these very same arguments in fact cause countless project failures. Surprisingly the rate of failure has hovered around the 70% mark for the same length of time. However, the problems and issues analysts write about are merely symptoms of much deeper problems in the field of software development. Where the majority of problems stem are from the underlying sociological basis for software development as whole. At the top-most level you have a host of bad management who barely understands both the technical and Human dynamics that go into making quality software. Most managers who believe they know a lot about such development are often extremely poor about understanding the necessities of the process from an overall process. Most technical managers, especially in the realm of business applications are nothing more than political hacks who are there to push projects through for the purposes of deadlines they have too often readily agreed to leaving quality as something you push on the developers after the fact. The next issue are the vendors themselves. Like most technical managers they are attempting to push software out the door to beat competition instead of deadlines, which are in themselves deadlines. The software is then full of bugs leaving software developers to develop timely work-arounds at crucial points in their projects. Then of course you have the avalanche of software that vendors keep on producing with new versions having newer technologies and methodologies included in every new release making it impossible for the software development community to develop a solid knowledge base whereby the issues and problems have ready answers. The last time such a knowledge base existed was with Windows 3.1. Finally, you have the massive ongoing dislocation that is constantly occurring to the software development community itself. With the rush to constantly save costs under the argument of increased competition, companies are abandoning any sense of commitment to the actual developer community themselves leaving them in massive states of insecurity. This in turn leads developers to concentrate on understanding technology for interviews and not good project development. Of course, you have the competitive spirit among software developers who are always attempting to work with the "latest and greatest" technologies but this again is a factor of the constant state of flux of the field that has jettisoned quality as a keystone to professional development for random knowledge. These contentions are basic generalities but they encompass the underlying factors for why software failure is a constant in the Information Technology field. After 30+ years of the same issues with the same reasons for them you would think that business would finally allow for some positive change in the software arena. However, this will never happen since business leaders tend to be very short-sighted and with very narrow priorities that always make up the realms which they make decisions within. Running off to India, China, and other such countries may have resolved some of the more short-term issues such as costs. However, in the long run such policies and endeavors have created a havoc in the industry that it cannot readily extricate itself from. The result will be merely more project failure and more articles like this one...

United StatesSteve Naidamast