Joseph Flahiff: Hey Michael.
Michael Mah: Hello Joseph, Good Morning.
Joseph Flahiff: I wanted to talk with you today a little bit about statistics [metrics] and information that is hard to find. I have been looking out there for data about Agile, real hard data about the productivity gains and the reduction in defects and those kind of things about when you move to using Agile and it’s really, really hard to find.
Michael Mah: Well it’s easy if you contact QSM or call me Michael Ma
Joseph Flahiff: And it is actually.
Michael Mah: Yes. That’s the inquiring minds want to know statement. And as the years have gone by, regardless of whether it’s Agile in the current state of the practice in the industry or before as object oriented programming before that it was reuse or what does it look like in different language environments, my firm has been tracking data for over 20 years. QSM associates, which is part of a larger organization worldwide, QSM with offices in Europe, in the States and the Far East have been capturing real statistics on software projects. Our database right now has about 10,000 completed systems and as data is feeding in from consulting engagements, where we are asked to help a company understand their patterns and also to be able to teach them how to measure on their own. We get data feeds [from these engagements] and this constant flow of information coming into our database, we add several hundreds if not thousand projects a year. And as we get new data that comes into our database, we actually take older data from the furthermost timeframe going back in history and put it on the shelf. All of this information stays current and contemporary.
Joseph Flahiff: That’s great, that’s great. So you’ve been collecting data then on Agile projects for some time now.
Michael Mah: Yes, Agile came from very lofty ambitions and when it first hit the radar a while back, the industry was reeling from projects that the norm was that they would blow by their deadline, miss their dates. Projects would overrun their budgets in the millions of dollars. When people said hold-the-line, shipped by this date and don’t spend anymore, functionality was cut because teams oftentimes were taking on far more than was doable inside their deadline or projects were shipped with known bugs. And the pain of that, especially if you are talking about a software that’s critical to an enterprise was excruciating and programmers were feeling blamed all the time. There’s got to be a better way than us going on a death march project which Ed Yordon (Death March (2nd Edition)) wrote very eloquently about in his books. Impossible deadlines, impossible scope, no hope and then at the end great you’re blamed for the “ failure.” But it was really a problem of the ecosystem. In some cases however, there were also very successful projects that did not suffer those kinds of outcomes. And they were waterfall as well. I took issue with some of the claims by the Agile community that Waterfall was bad and you were fat and heavy and overweight if you’re Waterfall. And if you were Agile, you were mean-and-mean and you know nimble and fast-moving. So the early claims were very lofty, it was pitted like a David against Goliath. The establishment at that time was Carnegie Mellon and the Software Engineering Institute and the Agilistas were making lofty promises but at that time they were mostly promises.
Joseph Flahiff: And have we seen the fruition of that in your data? What did the real data show and what are some of the challenges of running it.
Michael Mah: Oh great, let’s get right to the numbers.
Joseph Flahiff: Yeah, come on.
Michael Mah: You know I’ve given keynote speeches at conferences like Agile Development Practices that are software and I’ve also given lectures in Europe at the OOP conference in Munich, Germany. One of the titles of my talk is the Good, the Bad and the Ugly. What Agile data is showing us…
You will just have to listen to the podcast for the rest…