Whose software projects are canceled ?

The classic Standish Group 1994 Chaos report surveyed US companies engaged in the development of application software. Companies were ranked in size according to gross sales:


The outcome of software development projects were classified as Success ,Challenged ,or Canceled. Look at these grim statistics. The collective reporting of all companies in the Standish study show that the number of challenged programs or canceled programs both exceed the number of successful software development programs, as shown by the graphic.
Note: The sample size was 365 respondents and represented 8,380 applications.

What do project cancellations cost?

What do canceled software projects cost the software industry? The Chaos Survey reports these costs.

The graphic shows the distribution of software development cost for a $250 billion US software industry estimated for 1995.


Why are software projects canceled ?

In a September 2000 article by director Barry Boehm, the author recites statistics from the Chaos Survey and elaborates each. Software projects reported in the classic Standish Group 1995 Chaos survey, were terminated for these reasons :

In fact, as the Chaos Survey reports, the principal reasons for software development program Success, Challenge, or Cancellation were ranked in the top two or top five:

Software project termination does not necessarily equate to project failure, for example, in the case of lack of product demand or a viable market. However, the inability to develop a potentially successful product given lack of requirement definition or software resources is vexing in terms of loss of revenue and market to competition.

 Software process technologies, such as object-orientation and extreme programming, have been developed specifically to manage rapidly changing or vague requirements. The experienced wielding of these technologies minimizes the risk of project cancellation given incomplete and/or changing software requirements and lack of software resources, primary causes of cancellation for potentially successful projects in the studies referenced above.

References :

  1. "Project Termination Doesn't Equal Project Failure", Barry Boehm, UCSC, IEEE Computer, September 2000.
  2. The Chaos Report, Standish Group, 1994 (registration required)
  3. "Runaway Projects - Causes and Effects", Software World, vol. 26, no. 3.

A solution

The Standish Group reports that small companies do much better. Why not out-source your software development to a small software company? Software to Spec. designs and develops real-time, embedded software for a wide variety of microprocessor-based applications where fault-tolerance and reliability must be achieved in spite of user access, memory, and timing constraints. We have developed applications ranging from consumer electronics, networking and semiconductor devices, to commercial and scientific satellites and their simulators.

How do small vs large companies compare ?

Respondents to the Chaos Survey report that the small companies succeeded more frequently than did medium or large companies.

Small companies manage Challenge projects better than big companies too.

Finally, small companies performed better on Cancelled projects too.


In the most recent CHAOS Activity News Vol 2 : Issue 4 the Standish Group suggest  that companies take advantage of technical expertise and open source software development by "finding an experienced outsourced development team to get the project done. It's not just about saving money; it's about a quicker time to market."

How we do it ?

"Research at The Standish Group also indicates that smaller time frames, with delivery of software components early and often, will increase the success rate. Shorter time frames result in an iterative process of design, prototype, develop, test, and deploy small elements." This quote from the Standish report summarizes a spiral design process, a disciplined practice we at STS have followed since our foundation in the flight critical aerospace industry.

So what?  Why Software to Spec?

Software to Spec engineers practice the Software Engineering Institute's Personal Software Process.

"Because 70 percent of the cost of developing software is attributable to personnel costs, the skills, experience, and work habits of engineers largely determine the results of the software development process. This relationship of the engineer to the results of the development process is the premise on which PSP is based." (SEI) Our small company posture allows us to better control the software development process, from requirements definition, through design code and test. Specification change imposed by changing market forces can be quickly accommodated by changes to the UML model. Our careful application of software quality process early in the development phase allows us to complete projects 30% faster than conventional structured methodologies.

Software to Spec has the disciplined and experienced software engineers to deliver projects successfully. We are based in Europe and support clients in both Europe and North America. From our off-site location we can offer the combined software and hardware-diagnostic resources for developing your embedded and/or real-time system.

Software to Spec delivers:

Take the success test!

How does your software development program measure up? What is the success potential of your project? Take the Standish Group CAOS test for project success and see. A registration is required to gain access to this site..