Challenge a Recent Study Claiming “268% higher failure rates for Agile Software Projects”

The reason is that most of those were practising fake Agile. Real Agile software projects, with robust automated end-to-end UI regression testing, achieve nearly 100% success rates and save both time

I have come across posts on Reddit and LinkedIn referencing a study that claims “268% higher failure rates for Agile software projects,” but I have chosen to disregard them.

https://www.reddit.com/r/programming/comments/1d8r9xf/268_higher_failure_rates_for_agile_software/

Today, I saw an article on the reputable site The Register and felt compelled to provide some clarification based on my ~20 years of experience in agile development (clients and my own).

https://www.theregister.com/2024/06/05/agile_failure_rates/

About the Study

The original source I could trace to is this article on Engprax. The study “consisting of 600 UK and US software engineers finds projects adopting Agile Manifesto practices are 268% more likely to fail than those which do the opposite.”

The so-called findings of the study.

I think this study is biased because it was conducted for a new book, “Impact Engineering,” to promote so-called “Impact Engineering”. This is the main reason I have been ignoring this study.

The high failure rates in many so-called ‘Agile’ projects are largely due to the fact that most were practicing fake Agile. Genuine Agile software projects, which heavily emphasize End-to-end (E2E) UI test automation and Continuous Testing, achieve remarkably high success rates, saving both time and budget, with high quality, of course.

I anticipated the day when people would begin questioning Agile and consider reverting to Waterfall or adopting a different methodology. This shift is largely due to the widespread fake Agile projects over the past two decades. Throughout the years, I have written numerous articles warning about the risks associated with fake Agile.

• Be Aware of Processes Introduced by Fake Agile Consultants

• Why I don’t use Jira and Confluence at all for my software development?

• Estimating User Story Points is a Waste of Time

• Why Don’t I Use Defect Tracking? No Need, I do real Continuous Testing.

• Why Scaled Agile Framework® (SAFe®) is Wrong?

• User Story Card Clarified

• A Story of a Senior IT Manager Fears Real Agile, But Talks Highly About It

• A Story of Fake Agile Expert Selling “Release Early, Release Often”

• A Practical Advice to Shorten Daily Stand-up Meetings in Software Projects

Why do Fake Agile Software Projects (without solid E2E Test Automation) fail?