Designing, Learning and Tracking viral applications

Some of the notes of the 9.40 EST session with Jia Shen from Rock You! (Web 2.0 San Francisco). Apologies for any typos Design simple concepts Focus on virality and growth Accept the fact that channels Iterate Rapidly A/B/ Test Release Often. Rock you releases up to 20-30 new day There are no user experts Do user studies when you don't have web metrics Have a plan to maximize use of every channel

Development process:

Application types: Channel (Superwall, Huge Me) Content (Flixster) Dating (likeness, compare people) Games (Zombies, FluffFriends) Self Expression (Bumper Sticker)

Audience: Teens 14019 are 58% Very young, femails What's the total size of market What Percentage penetrations is goal Whats possible virallity of market? Average number of friends (average number of friends on Facebook about 150 on Myspace 300) Myspace vs. Ning. Vs. Google.


Develop your concept around high vonerting calls to action Simplicity Universal Social persuasion Novelty (art)

Viral Channels. How to grow your widget: News Feed Notifications Email Provile invites Non-user pages Profile action

Channel Focus 1 to 1  channels Map out several different flows to test -install to invite to interacting - Install to interacting to invite Balance relevance to throughput -channel vs. content applications.

Viral Loop Tune it - Superwall launched in 2 days tuned for 2 weeks, prview page, invite messaging, targeting Prmoote it You know the math works All about throughput to multiply Give it a boost Ad networks

Engagement Phase Saturated social circles (audiences) Critical mass makes features more useful tune for experience Build new features and keep users happy

Critical Mass You've achieved your gaol Pecentatge of Social Circle supernodes - 40% active 10% of that 4% of Market

Channels: Focus on 1-to-Many to get People Engaged - News Feed -Profile - NOtifications -Non USer Pages

Focus on 1-to-1 to re-engage current users - Email - Notifications

Hypothetical Viral numbers

Install flow x=5 (friends invited on average y=22% (acceptance rate for invites viral factor 5x0.22=1.1 VIRAL

Engagement flow\ - repeat users can generated additional virality - Engagement flow x=3 ( y= 10% (acceptance for invites viral cactor 3x0.10 No,3

Engagement Flow, notification channel x=6 y= 5% Viral factor 6x0.5%  Viral 0.3

What to track Request sent 7 Request CTR Nootifications send & Notification CTR Feed events & Feed CTR Adds/Removes And everything else - platform tracking is not 100% reliable

How to track Google Analytics Paid or home grown

Using metrics to drive product devisions A/B test entire user flows A/B test calls to action Base decision on metrics Trending graphs

Graph and compare acticity metrics ofver time and anlyze trends focus resorcues \on tuning the largest traffic dricvers all viral channels eventually decay So keep tunint to stay ahead of the curve

Session ended / Q and A