Skip to content

Going viral — not!

Sharad explains:

HIV/AIDS, like many other contagious diseases, exemplifies the common view of so-called viral propagation, growing from a few initial cases to millions through close person-to-person interactions. (Ironically, not all viruses in fact exhibit “viral” transmission patterns. For example, Hepatitis A often spreads through contaminated drinking water.[1]) By analogy to such biological epidemics, the diffusion of products and ideas is conventionally assumed to occur “virally” as well, as evidenced by prevailing theoretical frameworks (e.g., the cascade and threshold models) and an obsession in the marketing world for all things social. . . .

Despite hundreds of papers written about diffusion, there is surprisingly little work addressing this fundamental empirical question.

In a recent study, Duncan Watts, Dan Goldstein, and I [Goel] examined the adoption patterns of several different types of products diffusing over various online platforms — including Twitter, Facebook, and the Yahoo! IM network — comprising millions of individual adopters. . . . In all six domains the dominant diffusion event, accounting for between 70% to 95% of cascades, is the trivial one: an individual adopts the product in question and doesn’t convert any of their contacts. . . . only 1%-4% of diffusion trees extend beyond one degree. . . . cross the six domains only 1%-6% of adoptions take place more than one degree from a seed node, meaning that the vast majority of adoptions occur either without peer-to-peer influence or within one step of such an independent adopter. . . .

In all the examples we study, diffusion seems remarkably un-viral, rarely spreading far from an independent adopter. Our results thus call into question the dominant, epidemic-like models of diffusion, and also the value of viral marketing campaigns. On a positive note, this observation makes life a lot easier. Instead of needing to describe, predict, or trigger a complicated viral process, one can focus on the much easier case of adoptions that spread at most one hop before terminating. It turns out that diffusion is not nearly as messy as you might think.

I have nothing to add, except that in my experience these dudes know what they’re doing.


  1. Frank says:

    I also heard Duncan Watts talking about this in the audio version of his new book.

  2. In some way, this is similar to what I’ve heard about baby names going “viral.” I may be citing A matter of taste incorrectly, but I believe one if his major conclusions was that all parents are trying to be creative and innovative, and just happen to be creative and innovative in the same exact way. That’s how you wind up with “Jennifer” exploding in the 70s and 80s.

    And, if you think about it, the more people that some parents know who are giving their children the name X, the less likely they are to give their own child the name X (I assume). So if diffusion through networks were a prerequisite for a “viral” cascade, then viral-like patterns in baby names ought to be unexpected.

  3. Maximilian Hell says:

    “For the past several months, the [NY Times] R&D Lab has been working, quietly, on a time-based representation of how the Times’ news content is being shared in Twitter’s social space”

  4. Sylvia says:

    The analogy with ideas may not be so bad, if also in biological viruses only a small fraction of all new mutations become really contagious (via direct personal contact or otherwise). Does anyone have data on this?