The k-factor or viral coefficient measures how many new, secondary users an individual new user you acquire brings in over their lifetime.

For example, if every new user you acquire brings in, on average, 2 new users your viral coefficient would be 2.  The mechanism for bringing in new users can be anything: refer-a-friend, send an email, send a text message, send a Facebook request, etc.

If your viral coefficient is greater than 1 you will see logistic growth[1] and effectively reduce your cost of customer acquisition to zero.  This is why “viral growth” is so coveted in the consumer internet world.

In compartmental models of epidemiology[2], this is called the basic reproduction number.[3]

Note that the viral coefficient only tells you how large the next generation is as a function of the previous generation’s size, but not how much time there is between generations.

For example, imagine if you implemented a successful viral loop over email but waited 10 days before sending out emails to the addresses you collected.

It is hence inappropriate to compare viral coefficients, as typically measured, between applications.

[1] http://en.wikipedia.org/wiki/Log…
[2] http://en.wikipedia.org/wiki/Com…
[3] http://en.wikipedia.org/wiki/Bas…