This article has a bit of a 'just-so' vibe to me. What about a company like Uber? Uber doesn't have any virtuous loops, mounting loss, or cumulative benefits of use.<p>Now, I don't mean to argue against a straw man, I know that this article isn't claiming that <i>only</i> businesses with these features can be worth >1b - but exactly what is the ratio of companies north of 1b that have these properties? That is what really decides the predictive and analytic value of this model, not some cherry-picked stories that happen to adhere to it.<p>And also don't get me wrong, these things make intuitive sense to me as well. But i'd just like to see a more careful, rigorous analysis of these commonly accepted articles of faith in the startup world.