Constant Beta Is A Bit Like Hair Growing - No-One Notices
I'm sure you've had the comment, "Oooh, nice haircut" after a trim. The visit to the hairdressers has stopped people in their tracks, made them think about your hair as it was and compare it to how it is.
This, in IT terms, is the "big bang" approach and is used to put out new pieces of functionality/design in one hit. It generally requires a lot of project management, change management and communications and is inherently risky - all or nothing.
Of course sometimes the big bang approach can't be avoided:
However, many (most?) pieces of software do not need a complete visit to the hairdressers with such a dramatic change. They are like the everyday combing, curling and general coiffuring that we all do to make it a bit better.
And, like hair growing ever so slowly each day of "constant beta" it can be hard to track it's progress along the path of being more useful to users. In the web world a common approach it to perform A/B testing using the massive amounts of data being collected from the large user population. How to do this within organisations is both new and technically difficult (the tools and data just aren't there).
Back to "constant beta". When explaining the concept the question I use to highlight it in action is, "Can you remember how Google Search used to be when you first used it AND is it different now? But when did that 'upgrade' happen?". Google Search is a different beast but there was no "big bang" upgrade of functionality*, it has grown in usefulness before our very eyes.
But wait, Google has brushed off their 2001 index with its hairdo of the time letting us experience a time warp - we can really can see what "constant beta" actually feels like: http://www.google.com/search2001.html
* perhaps the unveiling of "universal search" back in May 2007 (Time article)
This, in IT terms, is the "big bang" approach and is used to put out new pieces of functionality/design in one hit. It generally requires a lot of project management, change management and communications and is inherently risky - all or nothing.
Of course sometimes the big bang approach can't be avoided:
- corporate re-brand, e.g. BNZ as of today
- introduction of new legislation, e.g. new NZ personal tax bands as of today
- introduction of brand new functionality, e.g. Google Moderator as of a few days ago (learn more)
However, many (most?) pieces of software do not need a complete visit to the hairdressers with such a dramatic change. They are like the everyday combing, curling and general coiffuring that we all do to make it a bit better.
And, like hair growing ever so slowly each day of "constant beta" it can be hard to track it's progress along the path of being more useful to users. In the web world a common approach it to perform A/B testing using the massive amounts of data being collected from the large user population. How to do this within organisations is both new and technically difficult (the tools and data just aren't there).
Back to "constant beta". When explaining the concept the question I use to highlight it in action is, "Can you remember how Google Search used to be when you first used it AND is it different now? But when did that 'upgrade' happen?". Google Search is a different beast but there was no "big bang" upgrade of functionality*, it has grown in usefulness before our very eyes.
But wait, Google has brushed off their 2001 index with its hairdo of the time letting us experience a time warp - we can really can see what "constant beta" actually feels like: http://www.google.com/search2001.html
* perhaps the unveiling of "universal search" back in May 2007 (Time article)
Comments
Post a Comment