Search Engine Marketing - Difference in Differences and Synthetic Controls

Political scientists or marketers will talk about a couple of cases in political science and marketing where you still want to be able to study causal impacts, but all you have is observational data. So what can you do in that case? That's what difference in differences and synthetic controls try to do. And let me use a particular paper as kind of rounding point here. Will have a link on linear digressions dot com.