The Standish Group conducts regular studies of project success rates which provide interesting insights into the project outcomes for software initiatives. Some of the data seems logical, but some of the findings took me by surprise. I would like to share some of these interesting data points for you to contemplate. Note that this data is aggregated from FY2011 through 2015, and consists of over 10,000 projects. You can click here to view the 2015 CHAOS Report.
Interesting Fact #1
On average, the percentage of projects that were successful between 2011 and 2015 was only 29%! If you told me that when I start a project, that there would be a 70% chance that it would fail, I would have probably not been happy!
Interesting Fact #2
Agile projects have almost four times the success rate as compared to Waterfall projects (39% vs. 11%), while Waterfall projects have three times the failure rate (29% vs. 9%). Is this a surprise to you? I had a feeling that Agile projects would be more successful generally, but did not expect the gap to be this wide.
Interesting Fact #3
Small projects have similar outcomes regardless of the method used. While this may not be a huge surprise, Agile projects have a 58% success rate compared to 44% for Waterfall, with failure rates relatively similar (4% for Agile vs. 11% for Waterfall). The gap is much larger for large projects, as you would expect.
Reflecting on this information, what are some conclusions that we can draw from this data?
Software projects are inherently challenging due to dynamic requirements and fluid technology. The requirements that are defined upfront have a high probability of being inaccurate or becoming obsolete as the operating environment is constantly evolving. This explains why the Waterfall approach has such as small rate of success. Even if the requirements were defined flawlessly, and the development efforts go perfectly, there’s always a high degree of likelihood that the end product no longer meets the needs of the customer the way that it was anticipated.
The data implies that we should seriously consider Agile methods for large software-centric projects since the typical outcomes are vastly different. This makes sense given that Agile enables faster feedback loops and earlier risk identification.
For training resources on Agile Project Management, see the following: