Quantifying Viral Loops With the K-Factor
Quantifying Viral Loops With the K-Factor
The viral loop we discussed in the previous post is something that has to be continuously measured and optimized. In order to do so, we must measure and quantify exactly how much virality we have, which can be done with “k-factor”. Looking back to that viral co-efficient we explained last time, the k-factor is the part of the formula that tells you have many new users have joined your service as a result of the viral activity of a single user in the period.
Determining where your k factor lies can be done using this mathematical formula:
C represents the number of cycles, time periods or generations; U(c) is the number of users at generation c and K is the viral coefficient aka “k-factor.” If you have values of K that are below 1, this means your viral growth is stagnating. When k is equal to 1, you have linear growth, and when k is greater than one….you have exponential growth and positive virality! That’s cause to celebrate.
In fact its not really a bad thing if you have a k-factor of 0.1 or 0.2 – this shows you have sharing and referral activity contributing to your acquisition and whilst its not stunning, its a great baseline to run new experiments on.
For example, if you do find users are sharing, then testing different methods of sharing can be explored:
- is sharing or inviting hidden or buried?
- can you tie it more into the user flow? For example Lyft prompts you after signing up to tell your friends.
- can you emphasize the sharing button? Can you tooltip/prompt/modal a user to sharing after a completed action in the App?
- have you targeted your power users with a push campaign incentives?
- For examples of what popular Apps do, check out our Referral Programs Guide
One thing to note is that the formula is very sensitive to small changes of K and so even a slight increase can have a huge impact. For example, a weekly growth factor of K=2 would be an almost unfathomable yearly growth rate of over 4.5 quadrillion. If you’re in the 1+ range, chances are you’re doing really well. In fact, a K-Factor of just 0.5 can nearly double your growth rate from your other marketing efforts. But even more important than your viral coefficient number is the Viral Cycle Time, which must be as short as possible in order to achieve profitable growth.
Now let’s look at a number of ways that k-factor growth is calculated:
- Reach: Once a share is made, its reach is the number of people that was shared to, or sent an invite. This depends on how easy it is to find and select all the people you would like to share to and the type of network it is made to. Sometimes there is a trade off with response rate: a share to social media will have a larger reach but lower response rate than direct invites sent via email or through paid search. Make sure the value proposition of your product is compelling enough that your customers will want to share it with others.
- Response rate: The proportion of people who click through on the share once they see it. This depends on the relevance and interest of the shared message, and the easier you make it to share the more effective the virality will be. Look at easy ways to implement sharing, such as through Facebook Connect integration or importing email contact lists with a pre-populated invite message.
- Engagement rate: The proportion of click through visitors who engage with the App and become a new user. This depends on how well virality is built into the App – your product designers and engineers should carefully orchestrate the first user experience to ensure engagement.
- Share rate: The proportion of users who share to others. The most viral products are ones that work only if shared. For example, early versions of Skype, Dropbox and Gmail were only useful if your friends were on them as well, and you had to get an invite (or invite others) in order to make this happen. Think about how to make your App social in a way that strongly incentivizes users to share it with their friends.
All of these growth methods can be individually tracked and optimized to make sure you are meeting your viral growth goals. We’ll discuss details and tips for optimizing each component in Part 3.