Airbnb Data Science team experience – Jeff Feng

Without experimentation, if we release a new feature and see the following graph, we can infer that with the introduction of a new feature the sales have increased. 

But if we could have done experiments, we may found that it is not true. In the absence of this new feature, the sales might have increased drastically, the reason for this can be anything. 

Controlled experiments are building block for decision making 

How useful is this feature 

Is it moving towards my and company goal 

How team is performing 

Characteristics about experimentation at Airbnb – different in different companies 

Data related tools and Infra 

They use tableau and superset extensively 

How to make sense of this data 


  Apache Superset 

It can analyze billions of rows in no time. 

Data Portal 

Airbnb has a data portal for all the data related work different teams/individual have done. They use the similar tech to google page rank, its kind of search engine. 

Data Education 

Data is at the core of everything. 

ML is everywhere in Airbnb. 

ML mapped to the every step of the process flow  


passionate about science and innovation. don’t settle, keep exploring…

passionate about science and innovation. don't settle, keep exploring...