Episode #14 - The Case for Waterfall, Six Sigma & Lean
In this episode of the MRX Lab podcast from FlexMR, we investigate three project management methodologies that are not agile. While agile research has quickly become part of the insights lexicon, there is much to be learnt from examining other techniques that have guided management principles over the past decades.
Throughout the course of this episode, we dive into the manufacturing roots of waterfall, the extreme ends of agile that comprise lean ideals, and the statistically driven Six Sigma process – evaluating the merits and drawbacks of each in turn.
Waterfall was born out of manufacturing processes and first applied to the field of software development in the 1970s. The methodology is deliberately sequential and liner. It requires a clear scope, and focuses on taking projects one step at a time. Typically these steps would be conception, initiation, analysis, design, construction, testing, implementation, and maintenance. Once when one stage is finished does the next begin.
It’s easy, looking back to identify the problem that agile was created to solve. It is a slow process, more concerned with delivering static, tangible outputs at each stage over speed. But this is also its advantage. A waterfall process is easy to understand, simple and easier to measure. It is fantastic for managing simple projects where scope is unlikely to change. Although it’s often regarded with disdain due to its lack of efficiency and dull sequencing, it is important not to underestimate how valuable these qualities can be – leading to more predictable timelines and results than alternatives like agile can offer. Prince2 is perhaps the most famous child of waterfall principles; developed by the UK government as an acronym of projects in Controlled Environments. It is a process-oriented system that emphasizes the need for justification in decision making.
Lean is a different approach; focusing on an MVP that is delivered at a much earlier stage and working from there. The MVP, or minimum viable product, is a small subset of the final product that forms only the least required for it to be usable. In product and software development, this means that the end result hits the market and is delivered much sooner – allowing for consumer feedback to help shape the eventual direction.
Six sigma posits that any problem can be solved through a five-stage cycle of: define, control, improve, analyse and measure. Some practitioners also factor in inputs and outputs, represented as stages of recognition and realization respectively. For researchers, this process should strike a similar chord that resonates with we conduct research itself. Six Sigma, however, offers a framework for looking at the processes through which we conduct this research – identifying areas of challenge and developing maintainable solutions.
Importantly, through analysis of these methods it becomes clear that agile research has a place in modern market research; but by focusing solely on the practice, researchers overlook a huge amount of knowledge that can be borrowed from the wider practice of management. We leave this on the table to our own detriment.