Viral Loop?

I’m reading the book, Viral Loop, a copy of which I was able to borrow from my public library.  While I have only read the Prologue, the Introduction, and the section on Facebook in Chapter 11, I have been fascinated by the stories of Hot or Not, Tupperware, and the early days of Facebook.  However, it appears to me that the author does not understand the difference between a positive feedback loop and a negative feedback loop.

In the Introduction (page 12), the author incorrectly characterizes positive feedback loops as “virtuous” and negative feedback loops as “vicious.”  For a negative feedback loop, the author provides the example of bad news causing a stock’s price to decline thereby causing more bad news thereby causing the stock’s price to decline further.  It is obvious that the author thinks this is a negative feedback loop because the stock’s price is heading in a negative direction.  In fact, (this stage of) the cycle described is a positive feedback loop, since the bad news causes more decline of the stock’s price and the decline of the stock’s price causes more bad news.  The changes increase each other in a vicious, destabilizing cycle.  I say “this stage of” the cycle because eventually the stock’s price will decline to a point where the number of buyers actually increases (good news).  So, the determination as to whether a system is experiencing positive (vicious, destabilizing) feedback or negative (virtuous, stabilizing) feedback depends on the time period during which the system is observed as well as on which inputs and outputs are observed.

A Ponzi scheme is a good example of a positive feedback loop.  The excellent results experienced by the early participants lead to more participants which leads to more excellent results.  Unfortunately, the continuation of such a scheme requires an infinite pool of potential participants.  There is no virtuous mechanism to stabilize the system and eventually there are too few new participants to pay off the earlier participants and the scheme collapses.  Viral expansion is not a good example of a positive feedback loop and in fact any system that exhibits a sigmoid (S-shaped) growth curve is not an example of a positive feedback loop.

In Chapter 11 (page 206), we learn that the founders of Facebook were “tacking up on their walls pictures of S curves representing the adoption rates on various campuses.”  Clearly the founders of Facebook were aware that the adoption rate was subject to a negative feedback loop.  The adoption rate increased at an increasing rate up to a point after which the adoption rate increased at a decreasing rate.  The point at which growth begins increasing at a decreasing rate is the inflection point.  On page 206, the author says, “The bigger the school, the longer it took to arrive at full inflection.  At Cornell a couple of weeks passed before exponential growth kicked in.”  I’m not sure the author knows what he means there.  I’m guessing that if you looked at the figures you’d find that exponential growth at Cornell began immediately and that it took a couple weeks until a limiting factor kicked in, resulting in an inflection point after which the growth rate was ameliorated.  Remember the example given in the Introduction:  doubling a penny every day for thirty days.  By the end of one week, you’d only have 64 cents, but the amount of money you have is (and has been since Day 1) growing exponentially.

I’m going to keep reading this book, but I’ll be wondering about the actual level of truth in everything the author says.

Comments are closed.