Predictive Analytics Innovation Summit, San Diego: Day 2 Highlights
Previous post
Tags: A/B Testing, CA, eBay, Facebook, IE Group, LinkedIn, Marketing, Predictive Analytics, San Diego
Highlights from the presentations by Predictive Analytics leaders from eBay, LinkedIn and Facebook on day 2 of Predictive Analytics Innovation Summit 2015 in San Diego.
Pages: 1 2
Andrew Ahn, Group Manager, Sales Solutions Marketing and Wenjing Zhang, Senior Manager, Business Analytics at LinkedIn delivered a talk titled "Growth Hacking with Predictive Analytics: What's after A/B Testing?". Andrew shared company mission and mentioned that they currently have more than 347M members. LinkedIn Sales Solution’s mission is to connect the world's buyers and sellers to build relationships. LinkedIn Sales Navigator is a product that was launched about a year ago helps sales professionals find the right people and companies, engage with relevant insights, and build trusted relationships.
Talking about A/B Testing, he displayed the two visual layouts for plan offers and shared results of the test. A/B testing is important because of the following reasons:
But one should also know what one doesn't get to know everything through testing. Using member demographic, his/her site engagement and social network one can know if the member would respond to the offer. LinkedIn has infused predictive analytics and data-mining techniques into its growth strategy to accelerate its growth trajectory.
Parsa Bakhtary, Games Product Analyst, Facebook talked about predicting the value of a social game install. In order to ensure that install ad campaigns are successful, a marketing team must have a reasonable approximation of the monetary value of each install from a person in a given demographic group, often called lifetime value (LTV) or revenue per install (RPI). Facebook has 375M monthly active gamers on mobile and web combined.
It was very difficult to come up with a generic predictive model that gives prediction counts over different game genres. Facebook computes cumulative 90-day revenue curves for weekly install cohorts of the top grossing Facebook canvas games. Parsa discussed the challenges in finding shapes and stability across various game genres. He concluded with some prediction techniques and applications to game ranking and the mobile space.
Related:
- De-risk business by evaluating multiple options
- Find small changes that lead to a big impact
But one should also know what one doesn't get to know everything through testing. Using member demographic, his/her site engagement and social network one can know if the member would respond to the offer. LinkedIn has infused predictive analytics and data-mining techniques into its growth strategy to accelerate its growth trajectory.
Parsa Bakhtary, Games Product Analyst, Facebook talked about predicting the value of a social game install. In order to ensure that install ad campaigns are successful, a marketing team must have a reasonable approximation of the monetary value of each install from a person in a given demographic group, often called lifetime value (LTV) or revenue per install (RPI). Facebook has 375M monthly active gamers on mobile and web combined.
It was very difficult to come up with a generic predictive model that gives prediction counts over different game genres. Facebook computes cumulative 90-day revenue curves for weekly install cohorts of the top grossing Facebook canvas games. Parsa discussed the challenges in finding shapes and stability across various game genres. He concluded with some prediction techniques and applications to game ranking and the mobile space.
Related:
Pages: 1 2