Rocket Internet: Attack of the online clones

A more precise description of their strategy would be: "go after big enough not-fully-claimed markets on native-preferring countries".

Airbnb has great execution, but they didn't specifically go after Germany early on, so Rocket popped up a clone and on China as well. Almost got bought from Airbnb at some point but greed prevailed and deal fell through.

Same with Groupon and probably others I can't recall.

No, their premise is that ideas as well as execution is cheap to replicate once it's already done by someone else in many of the cases. These people are actually copy pasting all the way down to HTML and CSS. All the design work and nitty gritty of how user interaction should look like is already done for them. They already know how it was bootstrapped, why it is successful and what users would expect from it. In other words they bypass most of the risky unknowns that causes large number of startup to fall. Backend is sure not replicable but given that they operate on smaller scales, most applications such as AirBnB would boil down to creating simple database and shuttling data back and forth which becomes college-level project that can be done quickly and thrown in market. I would not expect them to clone startups that is heavy on CS skills such as those requiring machine learning or computer vision or large amount of computing power. They would mostly go for startups where formulation of user experience is key and implementation is basically a relational database with UI doing simple read/writes.