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 and money.

Zhimin Zhan

--

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/

Table of Contents:
· About the Study
·
Why do Fake Agile Software Projects (without solid E2E Test Automation) fail?
1. “The benefits of specification”
2. “Agile causing delivery not to be on time”
3

--

--

Responses (4)