All that being said Siqi’s presentation was very insightful. Below is the slideshare deck as well as my major takeaways and notes on it.

Major takeaways:

• use metrics to yield actionable insight, make them easy to manage
• linear improvements can have non linear impact in space
• optimizing the viral loop can be the difference between a dieing app and a successful app
• virality can discount your user acquisition cost to zero
• you can build a scalable business with incentivised user acquisition if you manage other drivers that make up k effectively. K=infection rate x conversion rate
• when determining things to test think about what you can test for optimization in viral channels like requests, notifications, feed forms, emails, profile box, share link, email, app tab

Notes:

• Siqi’s 3 A’s of metrics

1. Actionable–built for action not vanity, all experimentation should help you make your app better and more profitable. There must be a framework in place for why you are testing something and what you hope to gain from it
2. Accessible–easy as possible to run experiments, tracking metrics as easy as not tracking them. Separate metrics from game infrastructure don’t want to re-do work with each new game
3. Auditable–sanity check with other data sources

• Why focus on metrics–
  • linear improvements lead to non-linear results
  • k=infection rate * conversion rate
  • k=user acquisition discount
  • k can discount user acquisition cost to zero
• Viral loop (see slide 13 in deck above)–each part of the viral loop has associated actions to impact user behavior and key metrics. Here are major ones:
  • visit–user visits driven by product quality, new user experience, social proof. Key metric activation rate
  • infection–users reached by average user. Things affecting this self expression, invite flows, retention, call to actions, game mechanics, and allocation tuning. Key metric infection rate
  • conversion–number of users reached by average user that actually converted. This can be impacted by copy writing, persuasion psychology/mechanics, demographic targeting. Key metric click through rate.

Are you an app developer? Do you agree with Siqi’s metrics and logic for viral loop activation? What do you test for optimizing the engagement loop?

Serious business CEO Siqi Chen optimizing the viral loop through metrics I've l