Anil Dash has an interesting commentary about software project cost and time overrun in the context of this New Architect story.

My take:

1. A certain amount of responsibility on how to buy software products and how to manage software projects do lie with the buyer. The kind of enterprise software referred to by 'The New Architect' Magazine is usually bought by medium to large enterprises.

For the last 5-6 years, there had been a huge gulf in the the salaries of the developers engaged in Internet and enterprise software development and in the salaries of developers engaged in government, fortune 500 et al. I am not trying to insinuate here that the later type of enterprises did not get bright people. But anecdotal evidence do suggest that the majority of sharp kids working in the OO / current generation of enterprise software / emerging technology areas worked for start ups / product companies. There had been (and probably still are) unequal distribution of talent in these areas. Many a times, enterprises taking software purchase / implementation decisions did not get quality advise from their IT departments.

2. Consulting companies chosen to advise enterprises have more often evaluated products based on self interest rather than on client interest. What I said on January 25 (my links are acting funny, so I have quoted from there) about technology consulting companies while on the subject of corruption in Anderson is still valid:

"Technology consulting companies are ..... badly compromised by their alliance partnerships. There are very few consulting companies that doesnt pitch (subtly or directly) products in which they have a vested interest to their clients. It has now come to a stage where many good product company will not look to ally itself with a technology consulting company if it does not bring in business to the product company. In many case, it would need to commit sizeable revenue to the product company to be accepted as a partner. As a result, most consulting companies today are becoming more hustler than consultants. Most of the clients who went for the big-buck e-business roduct implementations in the last few years simply did not need to go into that kind of expenses to achieve what they needed. They allowed themselves to be steered them towards the big ticket items by the consultants. The chicken is now coming home to roost."

3.Because of the exponential growth in tech sector there simply wasn't enough competent resources available to do all complext project justice. A well known practice among some consulting companies was of pitching for a project with the best team and switching it with a less experienced team later. A less savvy client will not know how to handle practices like these.

4. Sometimes, the technical team would deliberately steer companies in directions in which they want to have exposure rather than directions which are most practical/beneficial for the organization. Many of the development efforts that were undertaken in the last few years even in smaller start ups were simply started because some people wanted to have exposure to sexy technology. I know of at least one CTO in a web based service provider company (doing well I might add) who feels many of the EJBs that were developed adn deployed earlier was unnecessary, badly designed and added significant performance overhead. He is now looking to slowly rearchitect/reengineer his platform.

All these factors also contributed to the current mess in the IT sector