At Facebook, we routinely work with very large datasets to drive product decisions; we run models to help us attribute and quantify marketing efforts to sentiment and product usage.

We constantly run experiments: from simple A|B tests to advanced set ups involving network partition.

A network experiment is an experiment that takes into account the structure of the underlying network on which an experiment is being performed for the design (and perhaps analysis) of the experiment.  We suspect, or otherwise know, before even running our experiment that a user's reaction to a condition in the experiment may depend on other users' assignments to conditions and therefore want to incorporate that knowledge into our experimental procedures.

This dependence of your response on other users is often referred to as "network effects". 
In this talk, I will illustrate three recent products that we analyzed in the marketing analytics team and the challenges we encountered as well as some of the insights we gained.