One of the many things I try to make my customers realize, is that their web application is never done.
There is always something that can be made easier to use, pages that can be made to load faster, new features that can be implemented, existing features that can be made more powerful, or landing pages that can be optimized.
Pixar is a big inspiration to me, so it was pretty cool seeing that John Lasseter has the same approach regarding their movies
Our films don’t get finished, they just get released
When you release a movie, that’s it, no more changes (unless you later release a Directors Cut or Special Edition or whatever that only a fraction of people will see). It must be hard for perfectionists like the Pixar crew to know that there are scenes being shown that could have improved.
In the software business - at least in the web software business - we don’t have that constraint. We can launch a product and continuously improve it after the fact. And we should.
As Paul Graham puts it:
If your first version is so impressive that trolls don’t make fun of it, you waited too long to launch
Feed the trolls; launch your product now. Make it real, then iterate the crap out of it based on feedback from actual users.
Hence the "perpetual beta" label, touted by web pioneers like Google: Release early, often and in small iterations. Obviously, this is also a boon for the users, who generally will see their feedback taken into account much more immediate.
To me, working in such an environment actually makes it hard to imagine working any other way - 1 major release every one or two years with bug squashing and intense guessing about what features are needed to make the users want to both stick to your product and upgrade to the next, shiny version... jeez.
Oh, and nice to see one of my heroes, Paul Graham, cited btw. :)
From a business point of view, releasing as early as possible is also cost effective; for internal use this delivers value faster, and for external use can generate cash flow earlier.
Baskerville and Truex examined the subject in 1999: http://vingus.com/course%20work%20data%20files/IT%208125/Growing%20Systems%20in%20EMergent%20Organizations.pdf
To me, working in such an environment actually makes it hard to imagine working any other way - 1 major release every one or two years with bug squashing and intense guessing about what features are needed to make the users want to both stick to your product and upgrade to the next, shiny version… jeez.