I recall seeing an article or paper somewhere that said that x% of
software projects fail before they even start, because project
sponsors define the project features, schedule and resources, when
they should only be able to specify two, leaving one flexible. (For
example, if I were building a deck, I could specify that I want a 200
sq ft deck, made of redwood(features), and that I want it built by
next Saturday (schedule), and that I want to only spend $500 on it
(resources). This is an impossible project: I can specify any two, but
not the third ($500 deck by next Saturday, but it certainly won't be
200 sq ft)...)

Someone has done research on this, and specified the percent of
software projects fail before they start because of this problem. I
recall the percentage being over 50%, perhaps 66%.

**I need a reference to a paper or article, including author, page
number, periodical name, etc. that I can use in a white paper, because
I am being challenged on this point... (Internet sites accepted, if
they are credible, and indicate that research was done, not pulled out
of the air)

Hello brianpalmer-ga

I think I’ve found the report. It was done by the Standish Group in
1994 and is frequently quoted in articles and research papers.

http://www.standishgroup.com/sample_research/chaos_1994_1.php
The Chaos Report
The eye-catching statement that you remembered is included on the
first page of the report. “The Standish Group research shows a
staggering 31.1% of projects will be canceled before they ever get
completed.” The report then goes on to discuss the issues you cite in
your question. Their Web site shows that they have continued their
work in conducting this type of research and CHAOS Chronicles III will
be released in January 2003.

http://www.softwaremag.com/archive/2001feb/CollaborativeMgt.html
Collaboration: Development & Management Collaborating on Project
Success
This is a follow-up article from the Standish Group.

http://www.cio.com/archive/070101/secret_content.html
The Secret to Software Success
This article discusses the Standish Group’s longitudinal study and
gives current example of progress with agile development methods.

www.it-cortex.com/Stat_Failure_Rate.htm
Failure Rate: Statistics Over IT Projects Failure Rates
The IT Cortex consulting group provides statistics from five major
surveys – including the Standish report. These might be helpful for
further research for your white paper.

In case you need additional sources for research reports for your
project, check out the following.

http://research.sdmagazine.com/
Research Papers and Reports
Browse and search for white papers, case studies, webcasts and analyst
reports from the CMP Media magazines.

Research Strategy
"software project failure" research

Best wishes for success on your white paper.
czh

I think you pose a very interesting question.

You have to have a complete and detailed Project Mgmt Process/System.
When I got hired to where I work now. I had to create a brand new
process and I also created an intranet to help document our projects.

Our system is this:

1) Project Request - More details the better results we get
2) Project Plan - The person doing the work creates a plan of how to
execute request.
3) Project Review - Review the Plan with the Requestor and make sure
it's planned in the way to get what they want.
4) Start the Project
5) Testing Phase - Make sure it meets all technical/functional
demands.
6) Done!