There can only be one
Posted by Ian Holsman
I attended a interesting lecture about innovation by Iain Cockburn today where he was talking about standards, and how in any standard war, eventually only one will succeed.
standard examples he used were:
- VHS vs Betamax and
- OS/2 and Windows 95
where in both these cases the technically superior product lost. If this is true (and he seemed to have lots of maths and other economists to back him up) it is interesting when you look at linux vs NT, Blue Ray and HDDVD,.. eventually one of these will win out, thanks to network effects / virtuous circles et al.
In the software world (rails vs java vs django for example) this might be
- more apps written for it
- more books/blogs about it
- more developers using it
- more comitters support the framework
- more features in the framework
- leading to faster development times (leading back to more apps written with it)
Personally I think duopolies can exist one they have a ecosystem of complementors in place, and that some ‘wars’ will last for a long time to come thanks to the Deep pockets of the players in the game, and how all the players know the game well enough to know the rules.
The other interesting thing is that all the things you traditionaly think of being ‘wrong’ about how a company operates (as the end customer) is actually shown as a method of creating a dominating standard.
- Create FUD
- Signal your intent by burning your bridges behind you, forcing you to move forward
- Lie about your features/performance, you can fix it in the next release once your standard has won.
- move as fast as possible to build your market (making sure your profitable, not a dot-bomb)
- trumpet your wins, to create the impression that everyone is using your standard
other less objectionable methods
- create complementors, (killer apps or must-see content)
- low entry prices for the chicken and egg problem
- low switching costs
Going back to rails I don’t think it has created the circle yet, the apps aren’t there yet/mature enough. Once these are in place say in 6-12 months time it might be game over for other frameworks, but until then a ‘killer app’ in another framework will make it a interesting race.
another interesting point.. you don’t have to be using the dominant standard to make money. being on a popular standard means more (and harder) competition. It might just be easier to learn a less popular framework and consult in it as you could charge a higher price as there are less people with your expertise ;-)
Cockburn’s a good guy, I like him. He joined BU the year after I graduated, but a few friends were taking his classes, said he was great, so I went to hear him once. Now that I’ve gone through the MBA shtick with its competition strategy teachings, I appreciate Cockburn’s work a lot more.
I don’t think every field boils down to “only one”. In the field of programming languages, we still have stuff with us from 50 years ago. Something might dominate a niche, but for many fields where a significant technical advantage translates into time/money, I don’t think it will be possible for a ‘good enough’ framework to completely dominate. Java will probably maintain the high end for some time to come. PHP is not going to go away any time soon. Heck, Perl and Tcl haven’t been that popular lately and there are still lots of people using them.
I guess it really comes down to why you choose a framework.
If each framework can distinguish itself in a manner that it is a ‘natural’ in a given application space (eg publishing, or community forums) and provides alot more domain specific help in that area It could then segment the market sufficently that they don’t compete.