developeranalytics [dA] | blog

We spoke on the analytics panel at the SocialMedia Business School event last week in SF. It was a smaller, close-knit crowd, but there were some great discussions ranging from analytics to brand advertising. Check out this blog post for a summary of notes.

Just wanted to re-highlight a few key points we touched on in this panel:

• Virality (”k-factor) as the new metric: What is fundamentally different between social networks and the web? It’s the power of the social graph and friend relationships in itself. At dA, we measure the “k-factor” of apps. K-factor is the average number of friends invited by each user who already is an app user. So, a k-factor of 1 means that your app essentially is on a path to double (since each user is inviting on average another user).

• Moving the “Referral” stage forward: How do you maximize virality? By focusing on the referral stage as the primary driver for app growth.



At dA, we like to think about traditional “frameworks” in novel ways. In this case, what if you took the “Referral” stage and moved it forward — even before a user is activated? This is essentially what “forced invite” apps on Facebook do, and it’s why they “work.” By doing this, these apps maximize the number of the friends each user touches. Once this stage is maximized, each proceeding stage (retention, for instance) is maximized as well. The more friends that are referred, the more there are to be engaged.

Of course, forced invite apps are no longer permitted on Facebook. Even still, for the new era of apps, moving referral ahead of retention is still a highly effective technique.

• Importance of seed groups for viral growth: At dA, we are also app developers — purely to better improve our analytics product offerings. Why? We believe it’s critical to understand the pain points of developers as well as understand viral / engaging strategies used by applications in order to produce the best tools for developers to use. One case study we did recently was take two identical applications and seed them from two different user bases. The outcome? 20,000 vs 500 daily active users. Bottom line: your users matter.