3 ms·

Err...the following bothers me: "The ratio of converting invites to new users is your viral coefficient, k. If this is greater than one you will see self-susta

Err...the following bothers me: "The ratio of converting invites to new users is your viral coefficient, k. If this is greater than one you will see self-sustaining, viral growth.2 If the coefficient is less than one each user will bring in a fixed number of new users, but the application’s growth is still linear. " The rate of change of your userbase can be written as follows: rateOfChange = randomSignups - usersLosingInterest + currentUsers x viralCoefficient If viralCoefficient > 0 and currentUsers*viralCoefficient - usersLosingInterest > 0, you've got exponential growth, not linear. Of course this is a drastic simplification, as all of these coefficients depend on a ton of other factors that will not be constant over time. But the point is, each user doesn't have to pull a full new user in order to see exponential growth, you just have to have some positive viral coefficient...

jfarmer 18y ago
"Growth" here refers to the size of the installed user base, irrespective of retention. By "self-sustaining" I mean it requires no external input of users to continue to grow. Using your model, let's say usersLosingInterest = 0 and 0 < viralCoefficient < 1. How do you get exponential growth from this? Each new user brings in a finite number of additional users, a function of viralCoefficient. Call this number N. This increases your growth by a factor of N, but it's still dependent on some outside source of users. If I acquire M users then through my viral process I get an additional M*N users total. Another way to think about it: if your viral coefficient is less than one then the viral growth of each cohort dies out. I see nothing exponential here. Am I missing something? This point is mostly incidental to my two main points, though: outlining how the viral coefficient affects the cost of acquisition, and an argument for a hybrid strategy of accelerated but not viral growth. What do you think of those points?

jfarmer 18y ago

"Growth" here refers to the size of the installed user base, irrespective of retention. By "self-sustaining" I mean it requires no external input of users to continue to grow. Using your model, let's say usersLosingInterest = 0 and 0 < viralCoefficient < 1. How do you get exponential growth from this? Each new user brings in a finite number of additional users, a function of viralCoefficient. Call this number N. This increases your growth by a factor of N, but it's still dependent on some outside source of users. If I acquire M users then through my viral process I get an additional M*N users total. Another way to think about it: if your viral coefficient is less than one then the viral growth of each cohort dies out. I see nothing exponential here. Am I missing something? This point is mostly incidental to my two main points, though: outlining how the viral coefficient affects the cost of acquisition, and an argument for a hybrid strategy of accelerated but not viral growth. What do you think of those points?