Paul GrahamVerified account

Maybe growth in this case shouldn’t be measured in terms of the standard metrics

Or maybe there is a hidden objective function, like surveillance and control

I would argue that following growth will always discover local maxima and that's why there is always the opportunity of "disrupting" an industry by landing in another part of the curve. Following growth is nothing more than gradient ascent. Not that it's bad or wrong though.

Are these counterexamples really so rare? What about devices designed to break after the end of the warranty? They seem to be the norm rather than the exception.

FB's early strategy of limiting access to college students turned out to be genius

Network effects are rare?

social network growth should be designed to have layered approach....create initial pack that the network principles naturally aligns with. From that point localized should be planned to grow at global like onion rings...people of different scales added..

Isn’t this due to the fact that forums/social networks are open fields (similar to websites) where anyone can write about anything? Thus, you “follow growth” by focusing what gets traction inside them & push it so more people can see it.