Nice point that there is a little google lock-in, when people have learnt how to find things with google. It's not just the query syntax, but also that we've learnt what kinds of answers it gets back.
Thinking further: lock-in doesn't give you much competitive advantage in comp tech - but it does grant you a buffer against competitors. It buys you time, to match their improvement, or even improve on them. Therefore, in tech, I think competitive advantage should be measured in time * : our good image gives us 1 week; user inertia is 1 year; adapting to another interface is 3 weeks; our server farm speed is 6 months; PageRank is 2 months.
These times only come into play when a competitor offers something better in some way (if there's nothing better, then these timers aren't engaged). It's like a better product is a pressure or voltage differential, and the competitive advantage is the resistance.
Of course, when you're ahead is the time to grow your competitive advantage, even though it's not needed, so that you have it when it is needed. It's an investment. The parts of the model are: the users you have; the relative attractiveness of a competing product; and what stops your users from switching.
Actually, I think this is just one kind of competitive advantage, and it only covers existing users (not new ones). For a start up, it's not enough to just retain users, you need to get them in the first place. You're better off focusing on getting them (by increasing the relative attractiveness of your product) rather than stopping them from switching... but Warren Buffett is always going on about competitive advantage - what exactly does he mean?
