In relation to the article "Agile Manifesto co-author blasts failure rates report, talks up 'reimagining' project" of Richard Speed interviewing Jon Kern - whilst Engprax was not offered an opportunity to comment prior to publication, we provided The Register with the following right of reply once the article was published to address matters which we believe would mislead readers if left unchallenged - The Register has since pledged to update the article with a summary of the points below:-
Mr Jon Kern is quoted in the article as saying: "[The suggestion] that the Manifesto says: 'Yeah, forego clear requirements' – why would you want to do that? That's just silly…"
However, this rejects what is written in the Manifesto, what the reality of the implementation of Agile is and indeed what other authors of the Manifesto have previously said.
Dr Bertrand Meyer's book "Agile!", described as the "first exhaustive, unbiased review of agile", was notably critical of Agile's lack of upfront design. Indeed, in his 2018 paper "Making Sense of Agile Methods ", Dr Meyer goes as far as to say: "The project failures that I tend to see nowadays in my role as a consultant (or project rescuer) are often due to an application of this agile rule; we don't need no stinkin' requirements phase, we are agile, let's just produce user stories and implement them as we go. A sure way to disaster."
We note that another co-author of the Agile Manifesto has since backtracked on advocating that Agile was created to forego clear requirements following the publication of our research, despite there being a video of them on stage stating this as the aim when reciting the history of Agile.
In criticising the research, Mr Kern refers to "Waterfall" as a competing methodology, however, readers should ask themselves if Waterfall exists, where are the user groups or conferences of its proponents, even historically? Searching for a Waterfall conference online will lead you to two fake conferences created as a joke.
The answer lies in the fact that Waterfall is a strawman and was first described in 1970 by Wynn Royce as an example of a flawed process rather than a process that was seriously advocated. Waterfall has never really existed.
Furthermore, the article refers to criticism around the methodology and conclusions of the new research [N.B. here referring to the article's narrative rather than Mr Kern's comments] . The fieldwork collecting the data was conducted independently and the research agency was given no indication of any preference of results. Indeed, those who have read the ‘Impact Engineering' book will note one of the findings was contradictory to what the author expected. Earlier research found that 98% of business decision-makers in the UK and 96% in the USA agreed with the definitions of success and failure used to evaluate software projects.
An external tabulation of the results was also conducted under the supervision of another PhD-trained scientist which was consistent with the findings. It is also worth noting that the research agency conducting the fieldwork produced one of the most accurate predictions of seat numbers in the last UK General Election just two weeks ago. A 2021 review found the methodology used in this research to be consistent with those conducted using alternative methodologies when applied to software engineer populations. Furthermore, the research and book was independently reviewed and edited by yet another doctorate trained scientist with complimentary expertise to the subject matter.
In conducting this research, our staff faced multiple stalking and hacking attempts, seemingly aimed to shut down study of Agile, and the individuals responsible were identified in the eventuality that protective legal action became necessary. We found that those behind this conduct were universally also making untrue claims about the research and also had vested professional interest in Agile. None had any postgraduate-level scientific training or subject matter knowledge of the research methods used.
Interestingly, in the spirit of continuous improvement, we have identified a potential methodological change we'd like to experiment with in future to see if it has any impact on the research, though seemingly we have not found any third-party criticism which has identified this despite the scrutiny the research has been under.
More generally; a tertiary review published in a 2021 paper ‘"Best Practice" without Evidence – Agile Software Methodology as Example' since been debunked by academic research.
We are passionate about being challenged and we welcome further research, particularly that of a greater evidential value and that which challenges our work. In the past we have made updates to our public work in light of new contradictory information when the grounds for such updates would not have been publicly known hadn't we chosen to share them. However, we are concerned that personal harassment of those publishing research critical of Agile may lead to a silencing effect to researchers and those critical of Agile - that which is no doubt desired by extremists in the debate.
It speaks volumes that it takes a company whose bread-and-butter work includes standing up to powerful institutions and providing pro-bono work for whistleblowers facing retaliation to be courageous enough to dare criticise Agile.
