I wanted to talk about taking a really mathematical approach towards growing a startup you're creating. One of the slides in Dave McClure's presentation that really made sense to me in doing this and a method taught to me at Slide was this concept of the "viral growth factor." Mixpanel, our analytics service, can help you track this.

Lots of mainstream widget/application companies take this seriously as an early predictor of how a new application they are building is going to succeed. This obviously is not limited to application companies, this works for anyone that wants to see growth in visitor traffic.

So let's break it down, though Dave really did a great job, with an example.

An Example

With SuperPoke, a Facebook application by Slide, Inc, will have an X that is the number of users who poke their friends (thereby invite them to the application), their Y will be the number of users that a user pokes (rhere's a mass poke feature so 100s of users can be poked in one session), and their Z is the percentage of users who essentially accept the poke and go to the application. Z is interesting because those users can funnel back into X and recycle the process again creating near-exponential growth.

Ideally you want to make all three variables: X, Y, Z as high as possible but sometimes you can easily just focus on one which is why analytics here are so important

What's important?

Z is easy to figure out because it's going to be rather constant and easy to predict, 10% of the users invited may accept, this is just the click through rate. Analytics can help optimize your Z to get the highest rate of return but it's difficult to get it really high, such as 70%.

Y depends, with application companies Y is easy achievable by just defaulting all friends to be selected or a "mass select" all function and this can give a false representation of how "valuable" your application is, we'll talk about that later. Facebook did really well due to Hotmail email importing and thereby achieved a high Y for distribution.  Figure out what your Y is and come up with ways to increase it, word of mouth is difficult to come by. You have to be quite clever with Y by letting users "poke friends" or "share songs" which are better avenues than simply explicitly asking the user to invite their friends.

X is by far the hardest part of this equation and if your X is bad this whole thing falls apart. In some cases,  I would say getting users to invite their friends can directly correlate to how "valuable" or "interesting" your application really is. The best part of X is, you really control it, you can get this to be really high if you make something great. Elfyourself created something very entertaining and I am sure they had a lot of users inviting their friends because the application directly involved their friends in the movie.

I would say the most important one is the one you have the most control over, if you can augment Y to be a lot of users quickly then it's definitely worth it. Once you've got the equation close to 1, use analytics to optimize all parts of this equation to maximize your potential.

With SuperPoke, it becomes more obvious, X these days is the only variable they can focus on because they already have maxed out or optimized Y and Z so creating immense user-value is going to be priority to grow faster.

A common pitfall

A lot of widget/application companies went down this path: if you can achieve really high Y (mass inviting by default) you can easily achieve large growth but there is a problem. If your visitor retention is so low eventually you can fatigue users so much that your Z (acceptance percentage) can start to fall. This is what happened to lots of application companies. they invited 100s of users but the acceptance rate started to dwindle but when they had millions of users, those users rarely came back and eventually these companies hit a saturation point where their user-base declines everyday. Make sure your product has a high X (which can correlate to user value) or have a plan to attain it.

We encourage all of our clients at Mixpanel to take this approach to building their startup if your goal is visitor growth since we have the tools to help you track it. If you're new to Mixpanel, check us out!