An iterative process, whereby development ‘spirals’ to the ultimate product design
History of Spiral Development
The model shown below with regards to software development was first described by Barry Boehm in his 1986 paper “A Spiral Model of Software Development and Enhancement”, which was followed by another paper in 1988. These papers introduced a diagram that has been reproduced in many subsequent publications discussing the spiral model.
Also in 1986 Hirotaka Takeuchi and Ikujiro Nonaka wrote the “New New Product Development Game” and then in “The Knowledge Creating Company” described a form of “organisational knowledge creation, […] especially good at bringing about innovation continuously, incrementally and spirally”.
Then in 2005 Bob Cooper & Scott Edgett included spiral development as a component in the revised version of their Stage-Gate® process, as described in their book, ‘Lean, Rapid and Profitable New Product Development’.
Some Comments on Spiral Development
It is designed for programmes where requirements may change or are not clear, or where the technical challenge requires a number of iterations.
Risk-Driven Process Model
Anchor point milestones
Series of design-build-test-feedback loops
Get something in front of customer as soon as possible, get feedback then revise and add detail
It’s either a form of Agile or a series of Waterfalls – Take your pick
It is ‘Point-based’ rather than ‘Set-based’
Some Further Comments on Spiral Development
In fact ‘spiral’ development has been practised for ever. It is just that we used to call it ‘trial and error’ or just ‘development’, and not talk about it as a process, because it did not sound very scientific – it was just what everyone did.
We certainly did not talk about it once the quality gang came along with the idea of ‘Right First Time’, because it made one sound as if one did not know what one was doing, if one practiced trial and error. After all, trial and error could be described as ‘Wrong First Time, Right Maybe Second or Third Time if You Are Lucky’* – The clue is in the name trial and error.
Now it has a fancy name, and is back in fashion, we can shout about it and say we were right all along.
* Thomas Edison is said to have tried 6000 different organic materials in his search for a durable filament for his light bulb, and one survey found that the average number of design iterations before release was 4.
Elite’s Take on Spiral Development
Barry Boehm’s version of spiral development tries to steer a path between a Waterfall process and an Agile process by focussing on risk reduction. Bob Cooper’s version tends to focus on continually going back to the user with prototypes of some sort to optimise the product to meet the often unspoken user requirements to maximise the commercial success of the product (any kind of product, not just software). On the other hand Nonaka & Takeuchi are talking about the way knowledge is developed and transferred, rather than product development per se.
Elite’s view is that spirals of some sort or another are likely to happen in most product developments. Whether they are design iterations to overcome technical problems, or to improve manufacturability, or they are a means to elicit a better understanding of the customer’s requirements doesn’t really matter. The point is that they are there to develop knowledge and improving the product in some way.
Given that they are likely to happen whatever you do (at least the technical ones are, those that are intended to elicit customer requirements probably won’t happen unless you programme them in), the sensible course to take is to try to work out where the vulnerabilities are likely to be and allow enough time, resources and money to undertake a sensible number of iterations, rather than having them happen as unplanned surprises.
We are incorporating this thinking into the latest version of our recommended NPD process.
If you wpould like to discuss the applicability of spiral development in your NPD process, then please contact us.