| deutsch | english
Start | Sitemap | print | Search: 
SOPHIST > Requirements > Studies
Facts and figures

Studies – Why implement requirements engineering?

You’ve always wanted to know, what the advantages of implementing requirements engineering are? You need numbers, information and hard facts in order to convince your executives? We’ve gathered current, watertight evidence from both, highly theoretical and rather hands-on studies documenting why requirements engineering offers great benefits. These may supplement the experiences we’ve made and the commendations we’ve received and detailed in our trainings, project and books.

Chaos Report

Do you know its more current versions? The Standish Group, an independent marketing research institute, determined in 2000 that on an average, only a third of all IT projects are completed successfully. The data from 2002 corroborates this (34%). These numbers come from a census performed with American IT managers. Another striking number is out of the 2003 version of the report, which says that only about half of the features originally demanded are typically present in the final implementation of a software product.

Organizational & methodical aspects

Matthias Rauterberg discovered similar problems occurring in Swiss businesses. If you take a closer look at the organizational and methodical aspects of the main problems troubling Swiss IT projects, it quickly becomes clear that two thirds of these problems are RE-specific – meaning that if RE had been implemented in these projects, these problems wouldn’t be occurring. For Rauterberg, the top three obstacles are 1. the specification obstacle, 2. the communication obstacle and 3. the optimization obstacle. His article in the Journal of Industrial Ergonomics is well worth reading.

Cost Overrun according to Young

Investment into RE saves on overall project costs. That is what a study done by Ralph Young, published in his book “Effective Requirements Practices”, details. According to this study, projects that spend less than 5% of the overall budget on RE, end up costing between 2 to 3 times as much as originally estimated. If you spend more than 5%, you’ll be reducing costs dramatically.

Bug-fixing

You probably know the study Barry Boehm conducted about the amount of money it costs to correct an error in the various phases of software development. The interesting thing is that in a recent study (2000), Joseph Carr showed that these numbers are mostly still valid today.

Have a close look:

Barry Boehms - Relative cost to fix problem per phase during system development

 

Sources of particular interest:

  • Rauterberg, M.; Strohm, O.; Kirsch, Ch.: Benefits of user-oriented software development based on an iteractive cyclic process model for simultaneous engineering. In International Journal of Industrial Ergonomics Vol. 16, S. 391-410, 1995
  • Young, R.: Effective Requirements Practices, Boston, Addison-Wesley 2001
  • Hofmann, H.: Requirements Engineering. A situated discovery process. Wiesbaden, Gabler 2000
  • Carr, J.: Requirements engineering and management: the key to designing quality complex systems. In TQM Magazin, Vol. 12, Nr. 6, 2000, S. 400-407

Have you conducted similar studies in your organization? Inform us and thus others!