A Narrow Escape How We Avoided the Data Warehouse “Death March”.pptVIP

A Narrow Escape: How We Avoided the Data Warehouse “Death March” Chris Lawson What is a “Death March” ? Edward Yourdon, 1997, Death March: The Complete Software Developer’s Guide to Surviving “Mission Impossible Projects” Some “Encouraging” Words DW projects have a terrible reputation, often producing mad customers and a frustrated project team. Many seek to “jump ship.” Further “Words of Encouragement” Approx 50% of large DW projects “fail to meet the desired levels of success,” often failing on first or even second attempts.” CIO magazine, “The Middle Ground,” 1/15/99. Traditional Approach: Will it Work? For large information systems, development has traditionally been an extremely structured process. Many days are spent on up-front analysis, requirements analysis, design reviews, etc. The strategy for these types of projects is to invest much time early, when mistakes are cheaper to fix. Traditional Software Project For large government-funded projects, this is reasonable because requirements are fixed. Change in “Ground Rules” The strategy described above often does not work well with large, private-sector projects, such as Data Warehouses. The “ground rules” defining a successful project are different. For example, in contrast to government projects, Data Warehouse requirements are negotiable, elusive, and will probably change. Change in Ground Rules (cont’d) “Just when you are about to deliver, expect the rules to change – then change again.” Steve Tracy, Intelligent Enterprise, “Close the Loop,” March 27, 2001. Our Escape Route: Prototyping Early phases were very limited in scope. We deliberately followed a simpler development process. The Prototype approach allows early demonstration of a working model with some features. Key to Prototyping Benefits With scope limited to only six key reports and one data source, our early successes provided confidence for later stages. Another Benefit: Credibility Prototyping allows team