The Seinfeld Approach to Requirements
by Barbara Davis
Software industry stats clearly show there is an urgent need for dramatic and immediate improvement to the way we develop our products.
Only 16.1% of software and technical projects are successful, and that of the successful 16.1%, only 20% of the implemented features are used all the time and 40% are never used at all! That means that only
1.24% of all proposed features are actually implemented and used! Further, 2.48% of all proposed features are implemented and never used!
The single-most common reason for these alarming statistics is requirements. A lack of practice formalization and holistic view of the causes and effects of failure, have led to the shotgun approach to
requirements. Periodic successes with this approach have, in turn led Business Analysts to believe that the fault lies elsewhere in the process.
| Attachment | Size |
|---|---|
| The Seinfeld Approach to Requirements.pdf | 141.35 KB |
Argue about the statistics when lacking facts about content
Scott of course is right that most folks think most of their projects are successful; and rdavreq is right that our definitions and measures of success probably are flawed. Two articles available on www.gopromanagement.com you may find interesting analyzing issues with the CHAOS methods and results:
REAL CHAOS, Two Wrongs May Make a Right
This article was featured in the Dec. 18, 2008 issue of the IT Metrics and Productivity Journal http://www.compaid.com/
Despite questionable measures and analysis that misses the most common cause of project failures, the CHAOS Reports' conclusions may be correct.
Drowning in CHAOS?
This is the REAL CHAOS article as published in the February 2008 Software Test & Performance magazine.