Release Velocity and Risk

A frequent heated topic of discussion on the transition from waterfall to agile development models is the perceived major difference on the risk associated with the release velocity change.

If we consider that every new release brings a number of changes, and every change constitutes a risk, we can reasonably conclude the probability for a risk to be higher the more frequent we release.

In the 80s/90s there were many factors supporting this model, resulting in most of the hardware, software and processes designed at the time to have an explicit intention to prevent or aggregate changes.

As we entered the 21st century with major developments in both society and technology, it was critical to rethink how to cope with changes on software engineering, with a major milestone being the publication of the Agile Manifesto. Today it is widely recognized that avoiding changes in software is neither possible nor desired.

Does increasing the software release frequency on it's own increases software quality and reduces risk ? No, you need to follow a wider and consistent set of software development practices which enable change while minimizing risks. Something long enough for another article :)