Well, one thing’s for sure: Adam L. Penenberg didn’t take Networks from Cornell. He recently wrote an article exemplifying Andreessen’s “Ning”, which is basically a Content Management System that allows users to create social networks really easily (it’s related to Facebook like phpBB and bbBoard are to forums). A counter article then appeared in TechDirt that pointed out everything that Penenberg’s article did wrong. While Ning is a fairly interesting concept to think about using the power of networks, in this blog I’ll mainly consider Ning’s growth (which can be characterized with viral network effects like diffusion discussed recently in lecture).

Penenberg decides that the term “network effect” can’t describe the growth Ning is going through, and coins the phrase “viral expansion loop”. However, Ning’s growth obviously characterizes a network effect with tells like the power rule growth speed and diffusion spreading the software to its users after achieving a certain threshold of usefulness. He also claims this “alchemy” (no joke, he uses that word) is little known outside of Silicon Valley. However, TechDirt points out that this is clearly the tried and true formula of network effects – the very ones we’re studying (clear across the country from Silicon Valley).

However, possibly most insulting to those with knowledge of network effects is Penenberg’s implications of the ease of creating these “viral loops” (network effects). He states that the process of trying to create a network effect “almost guarantees” Google and Facebook-like success, completely ignoring the fact that starting something like that requires careful planning and network deployment. In class, we saw several graphs and noticed how careful placement of initial technology adopters could create (cluster-limited) cascades. However, from those same graphs one can easily see about 4 nodes that wouldn’t cause a cascade for every one that would. That argument is the basis of TechDirt’s counter article – that although Penenberg makes Ning’s growth seem like a secret success-guaranteeing formula, it really reduces to tried and true network effects which by no means guarantee success if implemented poorly.